Analysis Of Atmospheric Precipitable Water Vapor Change Based On MODIS Data

Precipitable Water Vapor plays a vital role in global climate change. It is of particular importance to study spatial and temporal changes in the precipitable water vapor on regional scales in order to predict rainfalls, weather disasters and monitor flood and drought conditions. In addition, precipitable water vapor, a greenhouse gas, accelerates the warming trend with other greenhouse gases. There are many different methods of detecting precipitable water vapor, the principles of different methods are summarized, the advantages and disadvantages of the methods are also compared. The algorithm for retrieving precipitable water vapor is introduced by several Moderate Resolution Imaging Spectroradimeter(MODIS) near-IR channels, the expression is deduced.Two-channel ratio weighting method and Three-channel ratio weighting method are applied for water vapor retrieval over Wuhan. The results show that three-channel ratio weighting method is more appropriate for Wuhan. A group of weighting factors derived from genetic algorithm(GA) and six other groups of different weighting factors are selected to calculate water vapor, the factors derived from GA is more precise, which is 0.4,0.25,0.35.Precipitable water vapor in recent years(2001~2013) over Wuhan are analyzed. Results indicate that the annual mean of precipitable water vapor has an obvious increasing trend and increases in summer, decreases in autumn. Seasonal fluctuation in Water vapor is analyzed, the results prove that fluctuation in autumn is wide, the standard deviation is 5.65 mm and it is small in winter, the standard deviation is 3.89 mm.Unary regression model is used to analyze the relationship between the precipitable water vapor and ground-based meteorological factors. It turns out that, precipitable water vapor correlates with atmospheric temperature positively, the correlation coefficient is 0.775, and correlates with atmospheric pressure negatively, the correlation coefficient is-0.688. The multiple regression model is applied to analyze precipitable water vapor, atmospheric temperature, atmospheric pressure and relative humidity. The multiple regression equation is used to estimate water vapor and the estimated value distribution is concentrated due to the linearity of multiple regression equation. The relationship between water vapor and Air Quality Index(AQI) is analyzed, the change of water vapor and AQI displays opposite tendency. Chengdu and Shanghai are selected to analyze the influence of the geographic location, the results show that water vapor over coastal areas increase faster than that over inland areas.